Reading Twin Sectors

Can clone CD read data CDs in AUDIO mode? Since data CDs contain a header and subheader fields after the Sync for each sector, programs which read the CDs in data mode will not read the twin sectors occuring on the CD – atleast when i tried it in CloneCD it gave me an error for both the twinsector and following sector on a test disc i made which contains 1 twin sector. However, reading the disc as an audio track – i used Easy CDDA Extractor v5.04 (www.poikosoft.com) – i was able to read the entire image. All the data was correct including both sectors of the twin sector data (and the sector following it).

I used a plextor 12/10/32S w/v1.05 firmware for the extraction with MAX reading speed.

Some drives refuse to read audio if TOC says “data” and vice-versa. Well, at the moment, I don’t understand why this worked…

If the twin sector was read correctly, then the size of the image should be 2352 bytes larger than what the TOC indicates.

CloneCD doesn’t report the error you mention here, but this might be drive-dependant.

But even in audio, shouldn’t the Subchannel give it away?

Maybe because the disc was read in block mode? eg. The drive reports ~ 24 sectors at once?

On another note: Is there any way to insert/modify sectors other than a Hex Editor? I can read them in CD Mage, but not add or modiy them.

@SDG: Not every sector contains timestamps in the Q channel! According to MMC4, at least 90% of the Q-subchannel shall contain time stamps (I hate words such as “shall” in specifications :a )

Maybe for audio, some drive only seeks once, and then read consecutively as long as consecutive sectors are read, without rechecking the timestamps?

The image was shorter than the original, but the audio program i used to read the disc in audio mode always seems to lop off the last 2 sectors of any data disc i read, no data appears to be contained in them (isnt it part of the leadout track?). Another oddity is that i cannot read the data contained in the sector by reading backwards. if anyone is interested i can provide a clone CD image of the twin sector to experiment with. i’m going to make a new image which will contain 2 twin sectors, the first twin sector set will contain 1 bad and 1 good sector, the second set will contain 1 garbage sector and 1 with the correct data.

Also, you mentioned the TOC, well when i first tried burning the image with cloneCD it reported that the .sub didnt match the .img file, so i edited the .sub and duplicated the info for the sector that contained the twin sector, does anyone know what the structure of the .sub is? i was able to observe the pattern to find the correct sector to modify, but im not certain what data i just duplicated :slight_smile: having done that, i could burn the image in cloneCD, with a prompt that stated the TOC didnt match the length of the img, and gave me three options for burning, i choose to adjust the TOC to match the size of the image.

00005DC0 00 00 00 00 00 00 00 00 00 00 00 00 41 01 01 00 03 25 00 00 05 25 20 61 00 00 00 00 00 00 00 00
00005DE0 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
00005E00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00

00005E20 00 00 00 00 00 00 00 00 00 00 00 00 41 01 01 00 03 25 00 00 05 25 20 61 00 00 00 00 00 00 00 00
00005E40 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
00005E60 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00

Another oddity is that i cannot read the data contained in the sector by reading backwards.

Of course not, you had two twin sectors adajcent to one another.

Try adding about 100(?) dummy sectors in between, and one at the end.

So, the pattern will go:

Twin Sector 1A
100 Dummy Sectors
Twin Sector 1B (With different data).
Sector 2.

Then seek: 1, 2, 1.

That should allow you to read the twin sector, hopefully.

Read this thread for more info on the properties of twin-sectors and some info on the way seek works for a cd-rom.

Subchannel analyzer

Hey, is there any tutuorial somewhere where I can learn this stuff? It looks weird.

Originally posted by CCDKing
Hey, is there any tutuorial somewhere where I can learn this stuff? It looks weird.

My site has a Tages tutorial, if that’s what you mean.

If not, what are you asking for a tutorial about?

@CCDKing: get ECMA-130.pdf.

It’s no tutorial, but if you are patient, you’ll find your way through it :slight_smile:

@SDG: Not every sector contains timestamps in the Q channel! According to MMC4, at least 90% of the Q-subchannel shall contain time stamps

Even in data?

Where’s it say this?

I spoke about audio, and since the track is suggested being read as audio, the audio assumptions about subchannel data and seeking would apply.

Yeah, but if you’re writing as data, won’t every sector have timestamps, even if it’s read as audio?

Or do you mean that it will only check 90% of the time?

The writer will write whatever you feed it with in RAW-DAO-96 mode! If you don’t include enough timestamps, or too many weird data in the q-subchannel, seeking might become unreliable, but the writer certainly won’t fix anything. Otherwise, RAW-DAO-96 would be rather useless :stuck_out_tongue:

I meant that maybe a single twin sector will be read “normally”, because one sector with “weird” subchannels might not be enough to make the drive wonder: It could be something else than timestamps…I somehow doubt that reading data as audio helps for larger twin-areas works.

> However, reading the disc as an audio track – i used Easy CDDA Extractor
> v5.04 (www.poikosoft.com) – i was able to read the entire image. All the
> data was correct including both sectors of the twin sector data (and the
> sector following it).

Well, reading the disc as audio in one row surely makes it less sensitive
to subcode tricks, but to be sure you should add around 100 twin sectors
and check it you can read them back.